ITEM METADATA RECORD
Title: Localization in Long Range Communication Networks Based on Machine Learning
Authors: Sallouha, Hazem ×
Pollin, Sofie #
Issue Date: 19-May-2016
Host Document: Proceedings of the 37th WIC pages:128-135
Conference: 6th joint WIC/IEEE SP Symposium on Information Theory and Signal Processing in the Benelux location:Louvain-la-Neuve, Belgium date:19-20 May 2016
Abstract: As the number of devices connected to internet is rapidly increasing, it is expected that by 2025, every device will have a wireless connection, hence leading to trillions of wirelessly connected devices. Therefore, internet of things (IoT) with long range, low power and low throughput (e.g., Sigfox and LoRa) are raising as a new paradigm enabling to connect those trillions of devices efficiently.
In such networks with low power and throughput, localization became more challenging. However, in most of IoT applications (e.g., asset tracking) we are interested in localizing the nodes within a certain area, rather than estimating the exact position with global positioning system (GPS) coordinates. Therefore, the problem can be simplified to estimate the node’s sector. In this paper we propose a localization mechanism based on machine learning and assuming that some nodes in the sector are integrated with a GPS. By using these GPS-nodes as a reference the network can learn the position of the other nodes. The results revealed that using the user back-end measurements (e.g., received signal strength (RSS), number of base stations and the end-to-end delay) nodes can be divided in sectors. Then, the GPS-node is used to define the coordinates of the sector. Moreover, a trade off between number of messages and localization accuracy is illustrated.
Publication status: published
KU Leuven publication type: IC
Appears in Collections:ESAT- TELEMIC, Telecommunications and Microwaves
× corresponding author
# (joint) last author

Files in This Item:
File Description Status SizeFormat
WIC.pdf Published 557KbAdobe PDFView/Open Request a copy

These files are only available to some KU Leuven Association staff members

 




All items in Lirias are protected by copyright, with all rights reserved.